Filgrastim prophylaxis in elderly cancer patients in the real-life setting: a French multicenter observational study, the TULIP study

Neutropenia is one of the most frequent limiting dose toxicities in cancer patients [1, 2]. It mainly depends on the chemotherapy (CT) regimen and can generate serious life-threatening complications. The incidence of febrile neutropenia (FN) varies from 10 to 57% depending on the chemotherapy protocols [1, 3].

Age is an independent risk factor regarding the development of chemotherapy-induced neutropenia (CIN) in solid and hematologic tumors [4]. According to EORTC’s [3, 5] and ASCO’s [6] guidelines, an age higher than or equal to 65 is an aggravating factor related to FN and must be taken into account when deciding on granulocyte colony-stimulating factors (G-CSF) treatment. In France, median age at the time of diagnosis is 68 years old for men and 65 years old for women and approximately 30% of patients are followed by onco-geriatrics [7]. The general condition associated with age is also a significant risk factor, the physiological age being more relevant than the chronological age to predict the risk of chemotherapy-induced neutropenia [8]. Neutropenic complications are not only more frequent among older patients but also more severe, entailing more numerous and longer hospitalizations as well as a higher mortality rate [9].

In order to prevent these complications, elderly patients are often treated with less aggressive CT protocols or with lower doses whereas (i) age is not a contraindication to the use of standard CT protocols: patients over 65 can profit from the same treatments as younger patients with comparable effectiveness [10]; (ii) controlled clinical trials have shown that the use of low-dose or shorter duration CT decreases the patients’ overall survival [11, 12]; (iii) G-CSF proved to be effective in reducing the incidence of FN, namely with elderly patients [13, 14].

Since elderly patients are excluded from most clinical trials [15], few studies are currently available concerning the management of CIN among this specific population of patients [16, 17]. No study under real conditions of medical practice evaluating the use of filgrastim has been conducted on cancer patients aged 65 or over. The optimum treatment duration with filgrastim and the optimum time for treatment initiation are also not clearly determined.

The aim of this national, observational, and multicenter study is to describe the modalities of use of Tevagrastim® in a real-life setting and also the profile of the patients treated among a cohort of patients aged 65 and over and undergoing anticancer chemotherapy in oncology and onco-hematology.

The TULIP study shows that filgrastim is mostly prescribed to elderly patients as primary prophylaxis (PP, 80%). These results are consistent with those of a French study (NEXT study), which reported the use of primary prophylaxis in 91.7% of the patients aged 70 years and older [18]. In contrast, primary prophylaxis was less common in a German study (HEXAFIL) conducted on a younger population (mean age, 59.4 ± 12.7 years), as 59.1% of the patients received secondary prophylaxis [19]. When filgrastim was used as PP, the treatment was predominantly initiated during the first CT cycle (79.1%); only 6.2% of the patients started to receive filgrastim during or after the fourth CT cycle; more particularly, 71.3% of patients with a high risk for CIN (≥ 20%) have initiated filgrastim during the first CT cycle. These results clearly highlight the implementation in common practice of the recommendations of the EORTC (original version written in 2006 and updated in 2010), namely the initiation of G-CSF primary prophylaxis within the first CT cycle for all patients presenting a risk of FN ≥ 20%. A Spanish retrospective study (LEARN study) conducted in 2003 reported a significantly lower proportion of patients treated with primary prophylaxis (45%) [20]. In contrast, a European study (MONITOR-GCSF study), recently conducted on 1447 patients in 12 countries [21], reported a proportion of primary prophylaxis (72.3%) comparable with that of the TULIP study (83.5%). For the patients who received filgrastim as secondary prophylaxis (SP), the treatment was generally initiated after the first CT cycle (75.5% of the cases), likely to address a neutropenic event which occurred during the previous close CT cycles. The median time to treatment initiation was 1 day longer for PP patients compared with SP patients (4 days after the onset of CT cycle vs. 3 days), and 3 days longer for hematologic malignancy compared with solid tumor (6 days vs. 3 days).

In nearly all cases, filgrastim was administered by subcutaneous injections at a daily dose of 0.5 MIU/kg. The treatment was considerably shorter than those usually adopted in randomized trials [22‐27]. Mean duration was 5.2 ± 1.9 days for patients with solid tumor and 5.5 ± 1.7 days for patients with hematologic malignancy. According to SmPC of Tevagrastim®, filgrastim treatment must continue after the expected date of the nadir and until the neutrophil counts have returned to normal to induce a lasting response. It is not recommended to discontinue the treatment prematurely before the expected date of the nadir [28]. After a chemotherapy for solid tumors, lymphoma, and lymphocytic leukemia, filgrastim treatment can last up to 14 days [28]. After induction and consolidation treatments for acute myeloid leukemia, the treatment may be significantly longer (up to 38 days) depending on the type, dose, and regimen of the cytotoxic chemotherapy [28]. Nonetheless, clinical studies showed that patients are often treated for shorter periods of time [19, 27, 29, 30]. In a retrospective study conducted from 1998 to 2002 in USA, the mean duration of filgrastim treatment was 6.5 ± 3.1 days for NHL, 6.1 ± 2.9 days for breast cancer, and 4.3 ± 3.1 days for lung cancer [27]. Another American retrospective study conducted from 2004 to 2008 reported treatment durations shorter than 6 days in 74% of patients [31]. In the HEXAFIL study, the median treatment duration was 4 to 5 days depending on the cycles, and the patients treated with PP received longer treatments than those treated with SP (5 days vs. 3 days for the first CT cycle) [19]. On the other hand, identical treatment durations (5 days) were reported in the MONITOR-GCSF study and the TULIP study [21], thus indicating a well-established practice for the prevention of CIN. It should be noted that treatment duration was reported in an overwhelming majority of the patients (less than 0.5% of missing data) in the TULIP study, which contributed to the establishment of reliable conclusions for this parameter.

CT dose reduction and CT delay are main concerns in patients presenting with high-risk neutropenia, especially in elderly patients. When patients develop severe neutropenia (grade 3 or 4) or febrile neutropenia, most of the time, CT dose is reduced or treatments are delayed [32, 33]. These practices have a significant impact on the success rate of the treatments, especially for curative chemotherapies aiming to extend survival or maintain the quality of life. Primary prophylaxis using G-CSF was clearly demonstrated to prevent the development of neutropenia and to reduce the rate of infection and infection-related mortality in adult cancer patients receiving chemotherapy [34, 35]. However, little data is currently available with respect to elderly patients because they are often excluded from randomized trials. Furthermore, randomized trials conducted on Tevagrastim® did not address directly the effects on modifications of the chemotherapy [36‐38]. In practice, the rates of CT dose reductions and CT delays due to neutropenic events in patients treated with short-acting G-CSF (filgrastim, lenograstim) vary considerably between studies (ranging from 1 to 46%) [18, 20, 39‐43]. In our study, less than 10% of the patients required CT dose reduction or CT delay due to a neutropenic event (5.5% and 8.1%, respectively). These rates were lower for the patients who received PP (4.8% and 7.1%) than for those treated with SP (9.2% and 13.3%). These findings confirm the efficacy of Tevagrastim® for the prevention of CIN in elderly patients in a real-life setting.

There are several limitations to our study. First, the enrolled patients included those with various cancer diagnoses, disease stages, and chemotherapy regimen. Our results therefore cannot necessarily be extrapolated to specific patient populations. Second, we could only determine the risk of CT-induced febrile neutropenia for less than half of the patients included in the analysis, in accordance with the risks defined by the EORTC. Although the risk of developing FN from chemotherapy is a key element in the decision algorithm developed for the prevention of CIN, there is currently no classification system comparing the risk levels of all different protocols of chemotherapy. Whereas the 2010 EORTC guidelines offers a classification for the most common protocols [3], many are not included yet in this classification, therefore leading to a large number of missing data as regards the subgroups analysis for the risk of CIN. Consequently, the description of this parameter remains unclear and the impact of filgrastim treatment modalities on the risk of FN may be difficult to interpret.

Although less robust than clinical trials from the methodological point of view, observational studies have the advantage of enabling “real-life” data collection, and thus to reflect the current clinical practice when a sufficiently large number of patients is included. The TULIP study was conducted on more than 1000 elderly cancer patients (aged 65 years or older). These patients are often excluded from clinical trials and, thus, seldom described in the literature [14]. Therefore, the TULIP study was able to establish the profile of an expanding category of patients as a result of the aging population. In France, more than 700,000 patients each year receive cancer treatments in hospital settings [7]. As of today, the incidence of malignancies after the age of 65 years has increased 11-fold compared with younger adults [44, 45] and nearly 30% of cancer patients are onco-geriatric patients [7]. Hence, it is essential to gain a better understanding of the conditions of use of filgrastim prophylaxis in these patients, which was the goal of the TULIP study.

As a conclusion, the French TULIP study is taking place in a broadly international reflection on cancer management of the elderly patients [44, 45]. Results provide a snapshot of real-life conditions of use of Tevagrastim® in elderly patients who receive prophylactic G-CSF medication: treatment is mainly initiated in primary prophylaxis during the first CT cycle (66%) and mean treatment duration per cycle is 5.3 ± 1.9 days. No difference was observed with regard to age or overall FN risk. Treatment duration was slightly longer when filgrastim was used in primary compared with secondary prophylaxis. The rates of CT delays and CT dose reductions were in the range of those reported previously for younger adult populations. These results highlight the efficacy and safety of prophylaxis with Tevagrastim® as common practice for elderly patients and support the data obtained from previous randomized trials on younger patient populations. Safety data were consistent with the known safety profile.